New brain biomarker identifies pain sensitivity and may predict chronic pain risk

By Dr. Priyom Bose, Ph.D.Reviewed by Danielle Ellis, B.Sc.Jan 28 2025

Slower brain wave patterns and reduced motor cortex activity linked to higher pain sensitivity and chronic pain risk

A recent JAMA Neurology study focussed on determining an effective sensorimotor cortical biomarker for pain based on two metrics, namely, PAF and CME.

Understanding the biomarkers for pain

Over the years, many studies have proposed different pain biomarkers, including neural oscillatory rhythms, proteins, metabolites, lipids, and neuroimaging markers of mechanistic/structural abnormalities. These biomarkers help clinicians diagnose, prevent, and treat patients with chronic pain.

Scientists experience multiple challenges in determining pain biomarkers, potentially associated with limited sample sizes that hinder the performance of full-scale analytical validation using a machine learning approach. The lack of clinically relevant pain models and the limited reproducibility of the existing ones have prevented the establishment of prominent pain biomarkers.

Previous studies have shown that the corticospinal signaling involved in the subsequent motor response and the neural oscillatory rhythms involved in processing nociceptive input are important factors in shaping the subjecting pain experience.

A previous study established a connection between PAF, CME, and pain. This study revealed that a slower PAF before pain onset and reduced CME during prolonged pain (depression) are associated with higher pain levels. In contrast, a faster PAF and increased CME (facilitation) are associated with lower pain.

An individual who experience extreme pain during the early stage of a prolonged pain episode, such as post-surgery, are at a higher risk of developing chronic pain in the future. If an individual develops slow PAF before an anticipated prolonged pain episode and/or CME depression during the acute stages of pain, it could indicate a transition to chronic pain.

About the study

The PREDICT trial focuses on identifying a prominent sensorimotor cortical biomarker signature for predicting pain sensitivity based on two metrics: PAF and CME. This trial used a nerve growth factor (NGF) pain model based on prolonged myofascial temporomandibular pain induced by intramuscular injection of NGF.

PAF has been defined as the dominant sensorimotor cortical oscillation in the alpha (8-12 Hz) range, while the CME has been defined as the efficacy with which signals are relayed from the primary motor cortex (M1) to peripheral muscles.

Pressure pain thresholds, PAF, and CME were measured on days 0 (baseline), 2, and 5. PAF was collected using a 5-minute eye-closed resting-state electroencephalography recording from 63 electrodes. Sensorimotor PAF was computed after determining the components in the signal that had a clear alpha peak (8-12 Hz) on frequency decomposition and a scalp topography suggestive of a sensorimotor source. Transcranial magnetic stimulation (TMS) mapping approach was used to measure CME.

The current study generated a map of the corticomotor representation of the masseter muscle. The corticomotor excitability was indexed as map volume. In electronic pain diaries, participants related their pain (0-10) during various activities, and this data was collected each day, i.e., from day 1 to day 30, at 10 AM and 7 PM.

Study findings

A total of 159 healthy participants, which included 70 females and 89 males, were enrolled in the PREDICT trial. However, 150 participants completed the trial, whose mean age was 25.1 years. The participants received intramuscular injections of NGF on days 0, 2, and 5. PAF and CME measures indicated good to excellent test-retest reliability across sessions.

The pain scores of the participants in the training and test set were classified as high and low pain sensitivity. Logistic regression indicated slower PAF and CME depression were associated with higher pain. Regression coefficients of −1.25 and −1.27 for PAF and CME, respectively, predicted pain sensitivity.

The locked logistic regression model indicated the optimal probability threshold for being classified as high pain sensitive was 0.40, with an associated sensitivity of 0.875 and specificity of 0.79. To be labeled as high pain sensitive, a 0.40 probability threshold was applied to the current data. This analysis underscored that a facilitator would need a PAF of less than 9.56, and a depressor would need a PAF of less than 10.57.

A visually slower PAF was found in individuals predicted to have higher pain sensitivity than those with low pain sensitivity. In the masseter motor maps, a reduced CME was found in individuals predicted to have high pain. In contrast, those predicted to have lower pain had elevated CME.

It must be noted that the combined PAF/CME signature outperformed each measure individually. The model with biomarkers including sensorimotor PAF, CME, sex, Pain Catastrophizing Scale (PCS) total score, and PCS helplessness score only outperformed the model including covariates.

To assess the reproducibility of the findings, the analysis was repeated using different PAF and CME calculation methods. Regardless of the methods, logistic regression was found to be the best or equal best–performing model when applied to the training validation set, with an area under the curve (AUC) varying from outstanding to acceptable to outstanding.

Conclusions

The study findings indicated that the PAF/CME biomarker signature could be positively used to predict the transition of pain from acute to chronic. This biomarker could help clinicians treat individuals with chronic pain.